
\begin{table}[ht]
\scriptsize
\caption{Results from the models used to estimate the effects in \Cref{fig:fg4}}\label{tab:fg4}
\begin{center}
\begin{tabular}{lcccccccc}
\hline\hline
 & \multicolumn{2}{c}{Democratic Appointees} & & \multicolumn{2}{c}{Republican Appointees} & & \multicolumn{2}{c}{All Appointees}\\\cline{2-3}\cline{5-6}\cline{8-9}
 & (1) & (2) &  & (3) & (4) &  & (5) & (6)\\
\hline
Nontraditional Appointees & 0.003 & --0.003 &  & --0.022** & 0.012* &  & --- & ---\\
 & (0.007) & (0.006) &  & (0.007) & (0.006) &  &  & \\
Republican Appointees & --- & --- &  & --- & --- &  & --0.016** & 0.009\\
 &  &  &  &  &  &  & (0.005) & (0.005)\\
\hline
Outcome & Stlmt. & Deft. &  & Stlmt. & Deft. &  & Stlmt. & Deft.\\
 &  & Wins &  &  & Wins &  &  & Wins\\
Cases & 131,502 & 131,502 &  & 133,387 & 133,387 &  & 291,243 & 291,243\\
Treatment Judges & 164 & 164 &  & 86 & 86 &  & 264 & 264\\
Control Judges & 131 & 131 &  & 164 & 164 &  & 304 & 304\\
Randomization Blocks & \checkmark & \checkmark &  & \checkmark & \checkmark &  & \checkmark & \checkmark\\
Min. Units/Trt. Arm & 5 & 5 &  & 5 & 5 &  & 5 & 5\\
Appt. Pres. Controls &  &  &  &  &  &  &  & \\
\hline\hline
\end{tabular}
\end{center}
{\scriptsize \textit{Notes:} All models include district-division-year fixed effects, which we refer to as our ``randomization blocks.'' We also use an adjustment proposed by Lin (2013) for all control variables (see main text). Standard errors are clustered by judge. Statistical significance is indicated by stars: * $p$ < 0.05, ** $p$ < 0.01 and *** $p$ < 0.001. }
\end{table}
